在電腦視覺影像資訊擷取的相關應用上，通常第一步都會對影像序列進行移動物體的偵測。將我們不在乎的背景影像濾除，保留完整的前景資訊，讓系統可以專注於移動物體後端的高階處理，像是追蹤、辨識、分類和動作判別，使得它們的效果與速度提升。在本篇論文中，首先利用自我組織類神經網路建立了一個具適應性的背景模型，它可以有效地處理動態背景、緩慢的光線變化、陰影投射和模型訓練的初始化問題。然而，背景濾除法所造成的保護色問題卻是非常的嚴重。因此，我們提出利用改良式的分水嶺演算法來取得影像的空間資訊，並將此資訊與背景濾除法所得到的初步偵測結果透過前景區域選擇演算法來做結合達到改善保護色問題的效果。從實驗結果中可以看到，利用前景區域選擇演算法的移動物體偵測結果明顯的優於單純的背景濾除法。

Detection of moving objects in video sequences is the first relevant step of information extraction in many computer vision applications. We filter out the background image which we don’t care and keep the complete foreground image. By doing this, it provides a focus of attention for tracking, recognition, classification, and activity analysis, making these later steps more efficient. In this thesis, we build an adaptive background model by self-organizing neural network which can handle scenes containing moving backgrounds, gradual illumination variations, has no bootstrapping limitation, can include the shadows casted by moving objects into the background model. However, background subtraction leads a serious camouflage problem. Due to this, we proposed a foreground region selection algorithm which combine the image space information and initial object mask generated from improved watershed algorithm and background subtraction respectively. We solve the camouflage problem effectively by the proposed algorithm. We can see the detection results of proposed algorithm are better than background subtraction from the experiments.

[1] S. Chien, S. Ma, L. Chen, “Efficient Moving Object Segmentation Algorithm Using Background Registration Technique”, IEEE Transaction Circuits And Systems For Video Technology, vol. 12, NO. 7, JULY 2002
[2] C. Stauffer, W.E.L. Grimson, “Adaptive background mixture models for real-time tracking,” Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition 2, pp. 246-252, 1999.
[3] K. Toyama, J. Krumm, B. Brumitt, B. Meyers, “Wallflower: principles and practice of background maintenance,” Proceedings of the IEEE International Conference on Computer Vision 1, pp. 255-261, 1999.
[4] K. Kim, T.H. Chalidabhongse, D. Harwood, L. Davis, “Real-time foreground-background segmentation using codebook model,” Real-Time Imaging 11 (3), pp. 172-185, 2005.
[5] L. Maddalena, A. Petrosino, “A self-organizing approach to background subtraction for visual surveillance applications,” IEEE Trans. Image Processing, vol.17, no.7, July 2008.
[6] 蘇木春, 張孝德, “機器學習：類神經網路、模糊系統以及基因演算法則,” 全華科技圖書股份有限公司, 2003.
[7] H. Liu, Z. Yu, H. Zha, Y. Zou, L. Zhang, “Robust human tracking based on multi-cue integration and mean-shift,” Pattern Recognition Letters,2008.

[8] R. B. Fisher, Change Detection in Color Images 1999 [Online]. Available:http://homepages.inf.ed.ac.uk/rbf/PAPERS/iccv99.pdf
[9] J. D. Foley, A. van Dam, S. K. Feiner, J. F. Hughes. Computer graphics : principles and practice. (2nd ed. in C), Addison-Wesley,1996.
[10] R. Cucchiara, M. Piccardi, and A. Prati, “Detecting moving objects, ghosts, and shadows in video streams,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 25, no. 10, pp. 1–6, Oct. 2003.
[11] S.-Y. Chien, S.-Y. Ma, and L.-G. Chen, ”Predictive watershed: A fast watershed algorithm for video segmentation” IEEE Trans. Circuits Syst. Video Technol., vol. 13, No.5 May 2003.
[12] Image Segmentation and Mathematical Morphology http://cmm.ensmp.fr/~beucher/wtshed.html
[13] L. Vincent and P. Soille, “Watersheds in digital spaces: an efficient algorithm based on immersion simulations,” IEEE Trans. Pattern Anal.Machine Intell., vol. 13, pp. 583–598, June 1991.
[14] D. Hagyard, M. Razaz, and P. Atkin, “Analysis of watershed algorithms for grayscale images,” in Proc. Int. Conf. Image Processing, pp. 41–44, 1996.